(Difference between revisions)
 Revision as of 11:57, 17 October 2013 (view source) (→Simulating evolution and then pooled sequencing)← Previous diff Current revision (13:33, 17 October 2013) (view source) (→Simulating evolution and then pooled sequencing) (2 intermediate revisions not shown.) Line 8: Line 8: ==Simulating evolution and then pooled sequencing== ==Simulating evolution and then pooled sequencing== - * '''We start with a population of N individuals with a certain SNP distribution $P(M)$ ~ $N(\mu, \Sigma)$.''' + * '''We start with a population of N individuals with a SNP distribution $P(M)$ ~ $N(\mu, \Sigma)$.''' - To do this we can use Dick Hudson's MS to lay down neutral variants (but the distribution really does not matter). + We can use Dick Hudson's MsSel (MS with selection) to generate a SNP distribution. + + + * '''We then evolve the populations for g generations with selection for SNP j''' + In MsSel, we can specify an allele frequency trajectory for a site. We can generate a trajectory based on a simple W-F simulation with selection. + + + Or, we may have to use [http://arxiv.org/abs/1310.3234 forqs]. - * '''We then evolve the populations for g generations with recombination''' - To do this we can use Kessner's [http://arxiv.org/abs/1310.3234 Forqs]. Forqs allows the user to specify the recombination rate file which might be useful in the future. * '''Simulate pooled sequencing''' * '''Simulate pooled sequencing''' Line 19: Line 24: (2) $f_i$ (frequency of the ith SNP) ~ $B(n, f_i^{true})$ (2) $f_i$ (frequency of the ith SNP) ~ $B(n, f_i^{true})$ - -

## Current revision

Analyzing pooled sequenced data with selection Main project page
Previous entry

## Simulating evolution and then pooled sequencing

• We start with a population of N individuals with a SNP distribution P(M) ~ N(μ,Σ).

We can use Dick Hudson's MsSel (MS with selection) to generate a SNP distribution.

• We then evolve the populations for g generations with selection for SNP j

In MsSel, we can specify an allele frequency trajectory for a site. We can generate a trajectory based on a simple W-F simulation with selection.

Or, we may have to use forqs.

• Simulate pooled sequencing

(1) draw a coverage (n) from Pois(λ) where $\lambda \approx 40-50$

(2) fi (frequency of the ith SNP) ~ $B(n, f_i^{true})$